As　intelligent　manufacturing　enters　a　new　era,　autonomous　mobile　robots　have　entered　a　stage　of　high-quality　development.　To　propose　3D　positioning　that　better　meets　the　needs　of　arbitrary　position　capture　in　the　manufacturing　industry,　we　need　to　focus　on　innovative　ideas　based　on　deep　learning.　Based　on　the　dynamic　evolution　of　RGB-D　camera　development,　the　theoretical　analysis　framework　of　3D　positioning　development　is　built　according　to　the　internal　logic　of　data-driven.　Researchers　explain　the　3D　positioning　development　mechanism　jointly　generated　by　depth　model　and　pose　estimation　related　to　RGB-D　camera　and　YoloX-Tiny.　In　addition,　we　will　continue　to　explore　the　possibility　of　switching　to　high-image-quality　development　from　the　perspective　of　changing　image　quality　and　experimentally　deriving　3D　positioning.　The　purpose　of　developing　this　technology　is　to　provide　the　desired　results　for　the　manufacturing　industry,　and　researchers　are　committed　to　continuously　improving　the　success　rate　and　efficiency　of　grasping.　Therefore,　we　take　a　series　of　measures,　including　object　identification　and　positioning　control　based　on　deep　information　quality　inner　loop,　constructing　the　interaction　and　coordination　mechanism　between　the　vision　system　and　robot,　and　establishing　the　evaluation　system　of　grasping　performance　and　user　satisfaction　to　realize　the　high-quality　development　of　technology.　It　will　boost　intelligent　manufacturing　and　meet　the　needs　of　the　manufacturing　industry.　©　2024　IEEE.